Cognitive Simulation: Intersection of Large Scale Machine Learning and Scientific Computing
Event Type
Machine Learning Day
AI/Machine Learning/Deep Learning
Big Data Analytics
Parallel Algorithms
TimeWednesday, June 19th4:30pm - 5pm CEST
LocationPanorama 3
DescriptionLarge-scale scientific endeavors often focus on improving predictive capabilities by challenging theory-driven simulations with experimental data. We'll describe our work at LLNL using advances in deep learning, computational workflows, and computer architectures to develop a cognitive simulation framework that is able to interface large scale machine learning with scientific simulation. We will present how this framework in the context of Inertial Confinement Fusion (ICF) simulation, focusing one the challenges of scalable deep learning.

We'll discuss necessary advances in machine learning architectures and methods to handle the challenges of ICF science, including rich, multimodal data (images, scalars, time series) and strong nonlinearities. These include advances in the scalability of our deep learning toolkit LBANN, an optimized asynchronous, GPU-Aware communication library, and a state-of-the-art scientific workflows. We'll also how the combination of high-performance NVLINK and the rich GPU architecture of Sierra enables us to train neural networks efficiently and begin to develop learned predictive models based on a massive data set.
Informatics Group Lead / Computer Scientist